Am J Physiol Heart Circ Physiol 297: H2068 –H2074, 2009.
First published October 9, 2009; doi:10.1152/ajpheart.00525.2009.
Generation of nitric oxide from nitrite by carbonic anhydrase: a possible link
between metabolic activity and vasodilation
Rasmus Aamand,1,2 Thomas Dalsgaard,3 Frank B. Jensen,4 Ulf Simonsen,3 Andreas Roepstorff,2
and Angela Fago1
1
Department of Biological Sciences, 2Center of Functionally Integrative Neuroscience, and 3Department of Pharmacology,
Aarhus University, Aarhus, Denmark; and 4Institute of Biology, University of Southern Denmark, Odense, Denmark
Submitted 10 June 2009; accepted in final form 6 October 2009
acetazolamide; dorzolamide; acidosis; neurovascular coupling
CARBONIC ANHYDRASE (CA) is a zinc-containing ubiquitous enzyme present in numerous isoforms (27, 42) catalyzing the
reversible hydration of the major end product of aerobic
metabolism, carbon dioxide (CO2), into carbonic acid, forming
bicarbonate (HCO⫺
3 ) and a proton:
CO 2 ⫹ H2O 7 H2CO3 7 HCO3⫺ ⫹ H⫹
(1)
Besides having a fundamental role in acid-base homeostasis
and in electrolyte balance (42), CA also effectively links blood
O2 delivery to metabolic CO2 production from tissues. In being
one of the fastest enzymes known (27), CA in red blood cells
is able to generate a local decrease in pH due to CO2 hydration
in the short transit time (⬍0.5 s) (47) of blood in the capillaries
of active tissues. This local acidosis decreases the O2 affinity of
the hemoglobin through the Bohr effect, so that more O2 is
unloaded to active tissues at a constant O2 tension (20, 30, 46).
Address for reprint requests and other correspondence: A. Fago, Dept. of
Biological Sciences, Universitetsparken Bldg. 1131, Aarhus Univ., DK-8000
Aarhus C, Denmark (e-mail: [email protected]).
H2068
Although numerous inhibitors of CA have been identified
(29, 42), CO2 and HCO⫺
3 are the only known substrates for this
enzyme. In view of the strong similarity between nitrite and
2
HCO⫺
3 , having the same trigonal planar sp orbital geometry
and almost identical pKa values of their respective acids (3.4
and 3.5, respectively) and given that nitrite is known to bind to
the active site of CA and to inhibit CO2 hydration (18), we
examined the possibility that CA could use nitrite as a substrate
to generate nitric oxide (NO), a potent vasodilator (17, 36).
A major part of NO generated by NO synthases (NOSs)
becomes oxidized to nitrite, which is present in tissues at low
micromolar levels (2). Although nitrite is a natural end product
of NO oxidation, recent work has shown that it may also
function as an important source of NO in mammalian tissues
(2, 45), particularly in the cardiovascular system (11). Several
metal-containing enzymes, including globins (4, 37, 39), xanthine oxidoreductase (32), and mitochondrial aldehyde oxidase
(26), have been now identified that may effectively catalyze the
one-electron reduction of nitrite to NO under hypoxic or anoxic
conditions, where NO bioavailability also increases as cytochrome c oxidase becomes more reduced (35). Although the
complex coupling between low O2 levels and nitrite conversion
to NO in tissues has been extensively investigated, it is not yet
fully understood how the end products of energy metabolism,
particularly protons, regulate NO generation from nitrite. Earlier studies have demonstrated the nitrite-dependent formation
of NO due to a pH decrease in the ischemic heart (48) and in
aorta vessels (33). Modin et al. (33) were the first to demonstrate in a seminal paper that physiological nitrite levels induce
NO-dependent aortic vasodilation, particularly at acidic pH
and independently of NOS activity. In this process, chemical
(nonenzymatic) acidification of nitrite to NO appears too slow
to play a major role at the low nitrite concentrations normally
found in vivo (3), which suggests the existence of a specific
catalyst. In this article we show that CA is able to readily
generate high levels of vasoactive NO from nitrite at normal O2
levels and in response to a decrease in pH. This novel physiological property of CA would contribute to regulate local
blood flow in response to an increase in tissue metabolic
activity before the O2 tension falls below critical levels.
MATERIALS AND METHODS
Materials. Dorzolamide (Trusopt) was obtained from Merck Sharp
& Dohme. Sodium nitrite (NaNO2), potassium nitrite (KNO2), purified CA (from bovine erythrocytes, containing the CAII isoform),
norepinephrine (NE), acetylcholine, asymmetric dimethylarginine,
ZnCl2, indomethacin, and acetazolamide were obtained from SigmaAldrich (St. Louis, MO). Purified CA was heat denatured at 100°C for
10 min. Apo-enzyme was prepared from purified CA as described
0363-6135/09 $8.00 Copyright © 2009 the American Physiological Society
http://www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on June 18, 2017
Aamand R, Dalsgaard T, Jensen FB, Simonsen U, Roepstorff
A, Fago A. Generation of nitric oxide from nitrite by carbonic
anhydrase: a possible link between metabolic activity and vasodilation. Am J Physiol Heart Circ Physiol 297: H2068 –H2074, 2009. First
published October 9, 2009; doi:10.1152/ajpheart.00525.2009.—In
catalyzing the reversible hydration of CO2 to bicarbonate and protons,
the ubiquitous enzyme carbonic anhydrase (CA) plays a crucial role in
CO2 transport, in acid-base balance, and in linking local acidosis to O2
unloading from hemoglobin. Considering the structural similarity between bicarbonate and nitrite, we hypothesized that CA uses nitrite as a
substrate to produce the potent vasodilator nitric oxide (NO) to increase
local blood flow to metabolically active tissues. Here we show that CA
readily reacts with nitrite to generate NO, particularly at low pH, and that
the NO produced in the reaction induces vasodilation in aortic rings. This
reaction occurs under normoxic and hypoxic conditions and in various
tissues at physiological levels of CA and nitrite. Furthermore, two
specific inhibitors of the CO2 hydration, dorzolamide and acetazolamide,
increase the CA-catalyzed production of vasoactive NO from nitrite. This
enhancing effect may explain the known vasodilating effects of these
drugs and indicates that CO2 and nitrite bind differently to the enzyme
active site. Kinetic analyses show a higher reaction rate at high pH,
suggesting that anionic nitrite participates more effectively in catalysis.
Taken together, our results reveal a novel nitrous anhydrase enzymatic
activity of CA that would function to link the in vivo main end products
of energy metabolism (CO2/H⫹) to the generation of vasoactive NO. The
CA-mediated NO production may be important to the correlation between blood flow and metabolic activity in tissues, as occurring for
instance in active areas of the brain.
GENERATION OF NO FROM NITRITE BY CARBONIC ANHYDRASE
Vasodilation experiments. Experiments were carried out using a
microvascular myograph (model no. 700 mo, Danish Myo Technology, Aarhus, Denmark) for isometric tension recordings. NE (0.020
␮M), NaNO2 (10 ␮M), purified CA from bovine erythrocytes (10
␮M), acetazolamide (100 ␮M), and dorzolamide (250 ␮M) were
applied as indicated in Fig. 4. PSS, containing (in mM) 119 NaCl, 25
NaHCO3, 4.7 KCl, 1.18 KH2PO4, 1.17 MgSO4, 1.6 CaCl2, 0.026
EDTA, and 10 glucose, was equilibrated with 5% CO2-21% O2-74%
N2, and mounting, normalization, and control of viability were performed as described previously (5, 34). All segments were incubated
with asymmetric dimethylarginine (300 ␮M) and indomethacin (3
␮M) for 30 min (40) to inhibit endothelial NOS and cyclooxygenase,
respectively. In selected experiments, aorta segments were incubated
for 30 min with the selective CA inhibitors acetazolamide (100 ␮M)
or dorzolamide (250 ␮M) alone or in combination with 1H-[1,2,4]
oxadiazolo[4,3-a]quinoxalin-1-one (ODQ, 3 ␮M) to inhibit soluble
guanylate cyclase (9). To induce a stable contraction of the arterial
segments, NE was added to the myograph chamber after equilibrating
with 1% O2-5% CO2-94% N2 for 5 min. None of the inhibitors used
influenced NE contraction. The active tension was defined as the level
of contraction after the addition of NE subtracted by the level of
contraction after the 30 min of equilibration in PSS after the normalization procedure. The tension after the addition of CA and nitrite was
expressed as a percentage of the active tension.
Statistical analyses. All data are presented as means ⫾ SE with a
significance level of P ⬍ 0.05, where n represents the number of
individual experiments. Data were analyzed using an unpaired t-test
by the SigmaPlot 11.0 software (Systat).
RESULTS
We used a NO-sensitive microelectrode to detect NO production in aerated phosphate buffer solutions following the
addition of KNO2 (100 ␮M) in the absence and presence of
purified CA (100 ␮M) (Fig. 1, A and B). We choose pH values
Fig. 1. Carbonic anhydrase (CA)-dependent nitric oxide (NO)
production from nitrite is increased by dorzolamide (Dorz) and
low pH. A and B: representative traces showing the increase in
NO production detected by a NO-sensitive electrode upon
additions (arrows) of KNO2 (100 ␮M, final concentration) and
dorzolamide (250 ␮M, final concentration) to CA solutions
(100 ␮M) at pH 7.2 (A) and pH 5.9 (B) at 37°C. C: maximal
NO concentration (means ⫾ SE; n, number of individual
experiments) measured upon addition of 100 ␮M KNO2 in the
absence (white bars) and presence (light gray bars) of CA and
in the presence of CA and dorzolamide (dark gray bars) at pH
7.2 and pH 5.9. The presence of CA alone increased maximal
NO (P ⫽ 0.003 at pH 7.2, and P ⫽ 0.001 at pH 5.9). Adding
dorzolamide resulted in a further increase in CA-dependent
NO production (P ⫽ 0.032 at pH 7.2, and P ⬍ 0.001 at pH
5.9). With CA present, the maximal production of NO from
100 ␮M KNO2 in the absence of dorzolamide (light gray bars)
was 13.6 ⫾ 2.4 nM at pH 7.2 and 95.8 ⫾ 15.7 nM at pH 5.9
(P ⬍ 0.001), whereas it was 83.2 ⫾ 33.1 nM at pH 7.2 and
205.1 ⫾ 23.1 nM at pH 5.9 (P ⫽ 0.011) in the presence of
dorzolamide (dark gray bars). D: control experiments showing
maximal NO concentration (means ⫾ SE; n, number of individual experiments) measured upon addition of 100 ␮M
KNO2 to phosphate buffer (pH 5.9) containing 100 ␮M of
ZnCl2, heat-denatured CA, or Apo-CA as indicated in the
absence (white and light gray bars) and presence (dark gray
bars) of dorzolamide. Note the enlarged y-axis scale. *P ⬍
0.05, significant difference.
AJP-Heart Circ Physiol • VOL
297 • DECEMBER 2009 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on June 18, 2017
(16), dialyzed against milli-Q (Millipore) water, and liophylized.
Adult male Wistar rats (10 –12 wk) were euthanized by decapitation
and subsequent exsanguination as approved by The Danish Ministry
of Justice (permission no. 2005/561⫺964). Tissues used for chemiluminescence were immediately excised, cut into small pieces, and
rinsed from blood in ice-cold PBS, homogenized, and stored at
⫺80°C. Arterial segments used for isometric tension recordings in
microvascular myographs were isolated from the thoracic part of the
aorta and were transferred to cold physiological saline solution (PSS,
see Vasodilation experiments).
NO production measured by a NO electrode. NO levels were
measured using a NO100 NO-sensitive electrode (Unisense, Aarhus,
Denmark) in 10 mM phosphate buffer at 37°C in an open chamber
with constant stirring. The response time of the electrode is ⬃10 s.
Dorzolamide, KNO2, and CA were applied as indicated in Fig. 1.
Solutions were prepared using milli-Q water. Traces were smoothed
by averaging with a running average spanning 10 s. The maximal
value for the trace obtained in this manner was used for further
analyses. The electrode was calibrated by injecting anaerobic milli-Q
water saturated with NO gas (2 mM) into anaerobic phosphate buffer
at 37°C.
Kinetic analyses. Kinetic traces obtained with the NO electrode
were analyzed by nonlinear least-squares fitting using the Table Curve
2D software (Systat, CA).
NO production measured by chemiluminescence. Experiments were
carried out in a reaction vessel containing 10 mM phosphate buffer
and KNO2 preequilibrated and purged with N2 at 37°C, using a
Sievers (Boulder, CO) Nitric Oxide Analyzer (NOA 280i) to detect
NO in the gas phase at a sampling rate of 4 s⫺1. Homogenized tissues
(2 mg/ml), dorzolamide (250 ␮M) and CA (10 ␮M), were applied as
indicated in Fig. 3. Traces were smoothed by a running average
spanning 10 s, and the maximal value obtained was used in further
analyses. Calibration was done by fully reducing the known amounts
of nitrite to NO with acidified potassium iodide, according to the
manufacturer’s instructions.
H2069
H2070
GENERATION OF NO FROM NITRITE BY CARBONIC ANHYDRASE
AJP-Heart Circ Physiol • VOL
Fig. 2. Kinetics of the reaction of nitrite with CA. A: normalized kinetic traces
of the reaction at pH 7.2 and pH 5.9 in the absence of dorzolamide, with trace
fitting according to a monoexponential function (gray line). Conditions were as
in Fig. 1. B: observed rates (means ⫾ SE; n, number of individual experiments)
calculated for each pH in the absence (white bars) and presence (gray bars) of
dorzolamide were not significantly different (P ⫽ 0.676 at pH 7.2, and P ⫽
0.393 at pH 5.9), and the data were therefore grouped according to pH
(cross-hatched bars). The reaction rates at the 2 pH values with and without
dorzolamide were significantly different (*P ⬍ 0.01).
reaction here studied (see Eq. 2) and, when purged in the gas
phase, completely homolyzes, yielding NO. As shown in
Fig. 3A, nitrite-dependent NO production was observed when
injecting purified CA at 10 ␮M in the purge vessel of the NO
analyzer containing nitrite (1, 5, and 10 ␮M) in phosphate
buffer at pH 7.2. This confirms the data of Fig. 1 and further
shows that the reaction also occurs under anaerobic conditions.
We then examined NO production following the injection of
rat tissues homogenates in the purge vessel containing 100 ␮M
nitrite at pH 5.9. Under these conditions, numerous pathways
of nitrite conversion to NO contribute to the NO detected. To
discriminate between NO produced from nitrite by tissuespecific CAs and by other pathways, we examined whether
dorzolamide had a significant effect on NO generation from
each tissue by performing chemiluminescence analyses in the
absence and in the presence of dorzolamide that specifically
reacts with CA to enhance NO production from nitrite, as
shown in Fig. 1. NO production from nitrite could be detected
in all tissues, in agreement with previous studies (8), with
blood, liver, and brain, showing a significant increase in the
NO generated with dorzolamide present (Fig. 3, B and C). In
particular, the high-NO signal observed with blood with and
without dorzolamide (Fig. 3, B and C) suggests that nitritedependent NO generation is in excess of the NO scavenging
capacity of the hemoglobin present, consistent with previous
297 • DECEMBER 2009 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on June 18, 2017
of 7.2 and 5.9 since they define the range from normal
intracellular pH to tissue pH after 5–10 min of ischemia (48),
respectively. As expected from its extremely low reaction rate
(3), little NO was generated in the uncatalyzed reaction,
whereas in the presence of CA, a significantly higher NO
production was readily observed when adding nitrite, particularly at low pH (Fig. 1). In an attempt to inhibit CA activity in
the NO-generating reaction, we added dorzolamide (250 ␮M),
a specific inhibitor of the CO2 hydration reaction catalyzed by
CA (42). This sulfonamide was chosen because of its high
solubility and because it did not interfere with the NO electrode, in contrast to the less water-soluble acetazolamide,
another specific inhibitor of CA (42). To our surprise, the
addition of dorzolamide increased the CA-catalyzed NO production to a significant extent (6-fold and 2-fold at pH 7.2 and
5.9, respectively, Fig. 1). The NO signal, which reflects the
balance between NO production and NO consumption, was
transient at pH 7.2 and prolonged at pH 5.9 (Fig. 1, A and B),
and the maximal NO signal was used in quantitative analyses
(Fig. 1C). The different profiles of the traces obtained at
different pH values suggest that in the course of the reaction,
intermediate or product species are consumed at high pH (Fig.
1A) or that nitrite is regenerated at low pH (Fig. 1B), although
the mechanism remains to be investigated. Heat-denatured or
Apo-CA enzyme (100 ␮M), Zn2⫹ (added as ZnCl2, 100 ␮M),
or dorzolamide alone (250 ␮M) did not cause any significant
increase in NO production from 100 ␮M nitrite (Fig. 1D).
Taken together, these data show that CA is able to react with
nitrite to generate NO and that higher levels of NO are
produced at low pH and in the presence of dorzolamide.
To determine the rates of NO production from nitrite in the
CA-catalyzed reaction, we analyzed the kinetic traces collected
using the NO electrode after adding nitrite to a final concentration
of 100 ␮M (Fig. 2A) to CA solutions in the absence and presence
of dorzolamide. Only the slowest response traces, indicating that
complete mixing of the nitrite injected had occurred in the reaction chamber, were analyzed. As the reaction was studied under
second-order conditions, i.e., at equivalent concentrations of enzyme and substrate (100 ␮M), kinetic NO traces could be well
fitted by a hyperbolic function (r2 ⫽ 0.981 ⫾ 0.024, n ⫽ 20). A
monoexponential function provided equally good fitting (r2 ⫽
0.975 ⫾ 0.084, n ⫽ 20) (Fig. 2A), suggesting that the concentration of free substrate nitrite was in excess of enzyme-bound nitrite
with the reaction taking place under pseudo first-order conditions.
This finding is consistent with the low affinity of nitrite reported
for CA (18). The observed rates were significantly affected by pH
but not dorzolamide (Fig. 2B), and the apparent rate constants
calculated from the values obtained in the monoexponential fitting
with or without dorzolamide were (means ⫾ SE) ⬎ 547 ⫾ 116
M⫺1 䡠s⫺1 at pH 7.2 and 84 ⫾ 42 M⫺1 䡠s⫺1 at pH 5.9 (Fig. 2B).
Because the half-time of the reaction at pH 7.2 (t1/2, ⬃13 s) was
equivalent to the response time of the electrode (⬃10 s), the
reaction at this pH occurs at rates faster than ⬃547 M⫺1 䡠s⫺1.
NO production at the low concentrations of enzyme and
nitrite found in animal tissues (2, 27) was measured using a
highly sensitive NO-chemiluminescence approach that detects
NO in the gas phase when purging a reaction vessel with pure
N2. In contrast to the experiments performed using the NO
electrode, where NO is detected in solution, chemiluminescence that detects NO in the gas phase also monitors the
production of N2O3 (3). N2O3 is formed in the course of the
GENERATION OF NO FROM NITRITE BY CARBONIC ANHYDRASE
H2071
observations (4). As dorzolamide alone has no effect on NO
generation (Fig. 1), this enhancing effect of dorzolamide suggests that a functional CA able to catalyze the conversion of
nitrite to NO is present in tissues (at least in blood, liver, and
brain), although the relative role of CA isoforms and of other
enzymes in the capability of different tissues to convert nitrite
to NO under physiological conditions remains to be solved.
To verify that under physiological conditions of low nitrite
(⬃10 ␮M) and high HCO⫺
3 (⬃25 mM with 5% CO2, pH 7.4)
that the NO produced by CA was vasoactive, we performed rat
aortic ring bioassays at 1% O2 and 5% CO2 (Fig. 4, A and B)
as previously described (5). A concentration of nitrite of 10
␮M, close to that existing in aorta vessels (2), is able to induce
a ⬃50% vasodilation under these conditions (5). The addition
of exogenous CA (10 ␮M) had only a slight effect on vasodilation in the absence of added nitrite, whereas the addition of
CA and nitrite caused a significantly larger increase in vessel
relaxation than that measured with either CA or nitrite alone
(Fig. 4C). The nitrite-dependent vasodilation of rat thoracic
aorta segments increased significantly in the presence of acetazolamide and dorzolamide (Fig. 4C). Conversely, vasodilation induced by the addition of CA and nitrite was significantly
reduced with ODQ and acetazolamide or dorzolamide present,
indicating that the relaxing effect observed is largely cGMP
dependent. These data indicate that CA reacts with physiological levels of nitrite to generate vasoactive NO via the activation of soluble guanylate cyclase.
DISCUSSION
Since CA is a ubiquitous enzyme, its ability to produce NO
from nitrite may be important in the control of nitrite-dependent signaling in general and in the control of local blood flow
in particular. Most enzymes known to function as nitrite
reductases are strictly dependent on severe hypoxic or anoxic
conditions (45) with the exception of xanthine oxidase in the
presence of NADH (45) and of hemoglobin, which generates
NO from nitrite most efficiently at PO2 close to its p50 (15, 21),
AJP-Heart Circ Physiol • VOL
although its physiological role has been questioned (5, 6, 19).
As shown here, CA is able to generate vasoactive NO from
nitrite at high rates and with no strict requirement for low O2.
This novel catalytic activity allows CA to translate an early
signal of increased metabolic activity (i.e., a fall in local pH
before O2 falls below critical levels) into an increase in tissue
O2 supply not only by favoring hemoglobin O2 unloading
through the Bohr effect but also by generating vasoactive NO.
Given that the active metal Zn2⫹ of CA does not sustain
redox activity, this enzyme most likely functions as a nitrous
anhydrase, with a reaction mechanism that is second order in
nitrite and equivalent to the reverse of the CO2 hydration
shown in Eq. 1:
2NO 2⫺ ⫹ 2H⫹ 7 2HNO2 7 H2O ⫹ N2O3
(2)
Where N2O3 rapidly homolyzes into NO and NO2 (3, 13),
N 2O3 7 NO ⫹ NO2
(3)
The reaction mechanism reported in Eqs. 2 and 3 corresponds to the acidic disproportionation of nitrite involved in
the acid-mediated and nitrite-dependent generation of NO
reported in previous studies (33, 48) and that is shown here to
be catalyzed by CA. Whereas purging by inert gases would
shift the above equilibria to the right due to the irreversible
homolysis of N2O3 in the gas phase (3), in solution N2O3 and
NO are in equilibrium (K ⫽ 2 ⫻ 10⫺5 M) (13) and the kinetics
shown in Fig. 2 better approximate the approach to equilibrium
achieved under physiological conditions.
The requirement for protons in the reaction in Eq. 2 explains
why more product is generated at low pH (Fig. 1). That the
reaction rate becomes faster at high pH (Fig. 2) suggests that
the anion nitrite (at high pH) participates more efficiently in
catalysis and further indicates that CA is not a nitrite reductase,
which increases reaction rate at low pH (15, 39). Interestingly,
the hydration of CO2 is also faster at a high pH, as the
Zn2⫹-bound hydroxide anion reacts more readily with CO2
than Zn2⫹-bound water (27).
297 • DECEMBER 2009 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on June 18, 2017
Fig. 3. Chemiluminescence analyses of NO produced from nitrite by purified CA and by rat tissues at 37°C at a 4 s⫺1 sampling rate. A: NO production from
addition of purified CA (10 ␮M, final concentration) to the reaction chamber containing physiological levels of nitrite (1, 5, and 10 ␮M KNO2) at pH 7.2.
Injection of CA caused a NO production that was linearly dependent on the nitrite concentration present in the chamber. A, inset: NO traces following CA
injections (indicated by arrows) at 3 different nitrite concentrations. B: representative NO traces following injections of rat whole blood lysates to the purge vessel
containing 100 ␮M KNO2 in 10 mM phosphate buffer (pH 5.9) in the absence (black) and in the presence (gray) of dorzolamide (250 ␮M). C: maximal NO
production (means ⫾ SE; n, number of individual experiments) following injections of rat tissue homogenates (2 mg/ml, final concentration) to the vessel
containing 100 ␮M KNO2 at pH 5.9 in the absence (white bars) and presence (gray bars) of dorzolamide (250 ␮M). Measurements with and without dorzolamide
were done in separate runs. All tissues added increased NO production from nitrite. In the presence of dorzolamide, NO production was enhanced in liver, whole
blood, and brain homogenates (P ⫽ 0.0150, P ⫽ 0.0170, and P ⫽ 0.0450, respectively). *P ⬍ 0.05, significant difference.
H2072
GENERATION OF NO FROM NITRITE BY CARBONIC ANHYDRASE
AJP-Heart Circ Physiol • VOL
297 • DECEMBER 2009 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on June 18, 2017
Fig. 4. Vasoactivity of NO produced in the reaction between nitrite and CA
and effects of acetazolamide, dorzolamide, and 1H-[1,2,4]oxadiazolo[4,3a]quinoxalin-1-one (ODQ). A and B: Representative traces showing the effects
of added CA (10 ␮M) and NaNO2 (10 ␮M) on norepinephrine (NE, 0.02
␮M)-contracted rat aortic segments in the absence (A) and presence (B) of
acetazolamide (100 ␮M) at 37°C and 1% O2-5% CO2-94% N2. Asymmetric
dimethylarginine (300 ␮M) and indomethacin (3 ␮M) were present to inhibit
endothelial NO synthase and cyclooxygenase, respectively. The traces illustrate the relaxing effect of physiological levels of nitrite in the absence
(continuous line) and presence (dashed line) of exogenous CA. C: degree (in
%) of relaxation observed (means ⫾ SE, n ⫽ 4 individual experiments) in the
absence (control) and presence of acetazolamide or dorzolamide. *P ⬍ 0.05
and **P ⬍ 0.01, significant difference. Acetazolamide and dorzolamide caused
a significant increase in nitrite-dependent relaxation both in the absence (P ⫽
0.0083 and P ⫽ 0.0222, respectively, light gray bars) and presence (P ⫽
0.0234 and P ⫽ 0.0346, respectively, dark gray bars) of added CA. ODQ
(white bars, n ⫽ 6) significantly decreased the degree of relaxation observed
in the presence of added CA and nitrite (dark gray bars) either in the presence
of acetazolamide (P ⬍ 0.001) or dorzolamide (P ⫽ 0.0018).
The yield of NO product but not its rate of formation
increases in the presence of dorzolamide and acetazolamide.
These two sulfonamide inhibitors of CA are commonly used
therapeutic drugs (42) and bind tightly to the active site Zn2⫹
to generate a tetrahedral complex that occupies the hydrophobic pocket where CO2 binds. That dorzolamide prevents the
hydration of CO2 but not the dehydration of nitrite suggests
that the two substrates bind to different groups in the active site
and may provide a first indication on the possible mode of
interaction between nitrite and CA. It is also possible that
dorzolamide and acetazolamide may increase the affinity for
the substrate nitrite by occupying nonproductive binding sites
on the enzyme. Interestingly, the active sites of Cu2⫹ bacterial
nitrite reductase and of Zn2⫹ mammalian CA are remarkably
similar (41), suggesting that similar amino acid residues may
be involved in the binding of nitrite in the two enzymes,
despite their different reaction mechanisms.
Earlier studies on CA have reported similar low binding
affinities for nitrite (KI ⫽ 63 mM) (18), HCO⫺
3 (KD ⫽ 70 mM)
(31), and CO2 (KD ⫽ 100 mM) (24, 27). Because of the low
affinity of CA for any of these substrates, only a small fraction
of the enzyme is bound to substrate under the conditions
investigated here and in vivo, which leaves a large enzyme
capacity available. Therefore, the proton-generating hydration
of CO2 and the proton-dependent generation of NO may occur
simultaneously in vivo, whereby end products of both aerobic
(CO2) and anaerobic (H⫹) metabolism would increase the
amount of NO formed from endogenous nitrite.
When local cellular activity suddenly increases, as it does
upon the activation of muscle or neuronal cells, the major
energy-generating pathway able to respond with comparable
rapidity is glycolysis, which initially contributes almost all
extra ATP produced. As a result, the local milieu will experience an initial temporary buildup of lactic acid and a concurrent decrease in pH, but no significant decrease in local PO2 due
to the release of O2 from the hemoglobin (Bohr effect). Thus a
local decrease in pH is the initial rapid signal of increased local
metabolic activity or of inadequate perfusion. The resultant
increase in the NO produced by CA from endogenous nitrite
under conditions of low pH (Fig. 1B) would provide a rapid
local vasodilation. The ability of CA to mediate a rapid
vasodilation is supported by high-rate constants for the conversion of nitrite into NO (⬎547 and ⬃84 M⫺1 䡠s⫺1 at pH 7.2
and 5.9, respectively, 37°C), which are much higher than those
for hemoglobin (⬃4.4 M⫺1 䡠s⫺1 at pH 7.0, 37°C) (15) and
myoglobin (⬃12.4 M⫺1 䡠s⫺1 at pH 7.4, 37°C) (39). Furthermore, as CO2 produced in the aerobic metabolism also decreases pH, NO production from CA could contribute to the
maintenance of vasoactive NO levels also during normal oxygenation. Our finding that dorzolamide and acetazolamide
increase CA-mediated production of vasoactive NO from nitrite (Fig. 4) may explain at least in part how systemic applications of acetazolamide increase blood flow in the brain (25),
lungs (14), and retina (44) independently of NOS activity (23).
These vasoactive effects of acetazolamide and dorzolamide are
largely dependent on the activity of soluble guanylate cyclase
(Fig. 4C), in agreement with previous results showing that
nitrite vasoactivity is inhibited by ODQ (5, 33). Whether the
effects of acetazolamide linked to NO generation are part of its
action in the treatment of acute mountain sickness (43) remains
to be investigated.
GENERATION OF NO FROM NITRITE BY CARBONIC ANHYDRASE
ACKNOWLEDGMENTS
We thank Lars Damgaard (Unisense) for analytical advice and assistance
with the NO electrode and Hans Gesser and Erik R. Swenson for comments on
the manuscript. A. Roepstorff is also at the Department of Anthropology,
Archaeology and Linguistics, Aarhus University.
GRANTS
This work was supported by grants from the Danish Natural Science
Research Council (to A. Fago and F. B. Jensen), the Lundbeck Foundation (to
A. Fago and T. Dalsgaard), the Danish National Research Foundation (to A.
Roepstorff), the Danish Medical Research Council (to U. Simonsen), and the
Danish Cardiovascular Research Academy (to T. Dalsgaard).
DISCLOSURES
No conflicts of interest are declared by the author(s).
REFERENCES
1. Berg JT, Ramanathan S, Gabrielli MG, Swenson ER. Carbonic anhydrase in mammalian vascular smooth muscle. J Histochem Cytochem 52:
1101–1106, 2004.
2. Bryan NS, Fernandez BO, Bauer SM, Garcia-Saura MF, Milsom AB,
Rassaf T, Maloney RE, Bharti A, Rodriguez J, Feelisch M. Nitrite is a
signaling molecule and regulator of gene expression in mammalian tissues. Nat Chem Biol 1: 290 –297, 2005.
AJP-Heart Circ Physiol • VOL
3. Butler AR, Ridd JH. Formation of nitric oxide from nitrous acid in
ischemic tissue and skin. Nitric Oxide 10: 20 –24, 2004.
4. Cosby K, Partovi KS, Crawford JH, Patel RP, Reiter CD, Martyr S,
Yang BK, Waclawiw MA, Zalos G, Xu X, Huang KT, Shields H,
Kim-Shapiro DB, Schechter AN, Cannon RO, Gladwin MT. Nitrite
reduction to nitric oxide by deoxyhemoglobin vasodilates the human
circulation. Nat Med 9: 1498 –1505, 2003.
5. Dalsgaard T, Simonsen U, Fago A. Nitrite-dependent vasodilation is
facilitated by hypoxia and is independent of known NO-generating nitrite
reductase activities. Am J Physiol Heart Circ Physiol 292: H3072–H3078,
2007.
6. Deem S, Min JH, Moulding JD, Eveland R, Swenson ER. Red blood
cells prevent inhibition of hypoxic pulmonary vasoconstriction by nitrite
in isolated, perfused rat lungs. Am J Physiol Heart Circ Physiol 292:
H963–H970, 2007.
7. Erusalimsky JD, Moncada S. Nitric oxide and mitochondrial signaling:
from physiology to pathophysiology. Arterioscler Thromb Vasc Biol 27:
2524 –2531, 2007.
8. Feelisch M, Fernandez BO, Bryan NS, Garcia-Saura MF, Bauer S,
Whitlock DR, Ford PC, Janero DR, Rodriguez J, Ashrafian H. Tissue
processing of nitrite in hypoxia: an intricate interplay of nitric oxidegenerating and -scavenging systems. J Biol Chem 283: 33927–33934,
2008.
9. Garthwaite J, Southam E, Boulton CL, Nielsen EB, Schmidt K, Mayer B.
Potent and selective inhibition of nitric oxide-sensitive guanylyl cyclase by
1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one. Mol Pharmacol 48: 184 –188,
1995.
10. Giacobini E. Localization of carbonic anhydrase in the nervous system.
Science 134: 1524 –1525, 1961.
11. Gladwin MT, Raat NJ, Shiva S, Dezfulian C, Hogg N, Kim-Shapiro
DB, Patel RP. Nitrite as a vascular endocrine nitric oxide reservoir that
contributes to hypoxic signaling, cytoprotection, and vasodilation. Am J
Physiol Heart Circ Physiol 291: H2026 –H2035, 2006.
12. Gordon GR, Choi HB, Rungta RL, Ellis-Davies GC, MacVicar BA.
Brain metabolism dictates the polarity of astrocyte control over arterioles.
Nature 456: 745–749, 2008.
13. Heinecke J, Ford PC. Mechanistic studies of nitrite reactions with
metalloproteins and models relevant to mammalian physiology. Coordination Chemistry Reviews. In press.
14. Hohne C, Pickerodt PA, Francis RC, Boemke W, Swenson ER.
Pulmonary vasodilation by acetazolamide during hypoxia is unrelated to
carbonic anhydrase inhibition. Am J Physiol Lung Cell Mol Physiol 292:
L178 –L184, 2007.
15. Huang Z, Shiva S, Kim-Shapiro DB, Patel RP, Ringwood LA, Irby
CE, Huang KT, Ho C, Hogg N, Schechter AN, Gladwin MT. Enzymatic function of hemoglobin as a nitrite reductase that produces NO
under allosteric control. J Clin Invest 115: 2099 –2107, 2005.
16. Hunt JB, Rhee MJ, Storm CB. A rapid and convenient preparation of
apocarbonic anhydrase. Anal Biochem 79: 614 – 617, 1977.
17. Ignarro LJ, Buga GM, Wood KS, Byrns RE, Chaudhuri G. Endothelium-derived relaxing factor produced and released from artery and vein is
nitric oxide. Proc Natl Acad Sci USA 84: 9265–9269, 1987.
18. Innocenti A, Zimmerman S, Ferry JG, Scozzafava A, Supuran CT.
Carbonic anhydrase inhibitors. Inhibition of the beta-class enzyme from
the methanoarchaeon Methanobacterium thermoautotrophicum (Cab) with
anions. Bioorg Med Chem Lett 14: 4563– 4567, 2004.
19. Isbell TS, Gladwin MT, Patel RP. Hemoglobin oxygen fractional saturation regulates nitrite-dependent vasodilation of aortic ring bioassays.
Am J Physiol Heart Circ Physiol 293: H2565–H2572, 2007.
20. Jensen FB. Red blood cell pH, the Bohr effect, and other oxygenationlinked phenomena in blood O2 and CO2 transport. Acta Physiol Scand 182:
215–227, 2004.
21. Jensen FB. The role of nitrite in nitric oxide homeostasis: a comparative
perspective. Biochim Biophys Acta 1787: 841– 848, 2009.
23. Kiss B, Dallinger S, Findl O, Rainer G, Eichler HG, Schmetterer L.
Acetazolamide-induced cerebral and ocular vasodilation in humans is
independent of nitric oxide. Am J Physiol Regul Integr Comp Physiol 276:
R1661–R1667, 1999.
24. Krebs JF, Rana F, Dluhy RA, Fierke CA. Kinetic and spectroscopic
studies of hydrophilic amino acid substitutions in the hydrophobic pocket
of human carbonic anhydrase II. Biochemistry 32: 4496 – 4505, 1993.
25. Laux BE, Raichle ME. The effect of acetazolamide on cerebral blood
flow and oxygen utilization in the rhesus monkey. J Clin Invest 62:
585–592, 1978.
297 • DECEMBER 2009 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on June 18, 2017
The finding of different production of NO from nitrite in
different tissues reflects the expression of tissue-specific CA
isoforms with different catalytic activity and sensitivity to
dorzolamide and the activation of other nitrite-reducing pathways under anaerobic conditions. The NO increase observed in
the presence of dorzolamide (Fig. 3C) indicates that the CAdependent pathway of generation of NO from nitrite may be
particularly active in blood, but also in liver and brain. Furthermore, both the CAI and CAII isoforms are present in the
vessel wall (1) and may participate to intrinsic mechanisms of
nitrite-induced vessel relaxation (5). CA is expressed in a wide
range of cell types, including capillary endothelial cells, glial
cells, and erythrocytes, and is either soluble, intracellular,
extracellular, or membrane bound (10, 27, 38). The most active
isoform CAII has a wide tissue distribution and is also the
second most abundant protein in red blood cells, which is
consistent with the large effect of dorzolamide on NO production shown in Fig. 3, B and C. Less CA is present in the brain,
in accordance with a reduced effect of dorzolamide (Fig. 3C).
The membrane-bound CAIV isoform is expressed mostly in
lung capillaries and kidney epithelial cells and is less inhibited
by dorzolamide, which might also have reduced the increase in
NO formation due to dorzolamide (Fig. 3C).
Because of its localization in blood circulation and neuronal
tissue, the NO-generating activity of CA is potentially of key
importance to the still poorly understood neurovascular coupling (28), i.e., the correlation between neuronal activity and
vasodilation used in analyses of neuronal processing by blood
O2 level-dependent functional magnetic resonance imaging
(BOLD-fMRI). Specifically, astrocytic glycolysis has recently
been found to be a determining factor in the regulation of local
blood flow in response to neuronal activity in the brain (12). As
NO is involved in numerous biological processes besides
vasodilation, including mitochondrial O2 consumption and
cytoprotection (7, 11), the novel NO-generating function of CA
reported here widely extends the physiological role of this
ubiquitous enzyme and the potential therapeutic applications of
dorzolamide and acetazolamide.
H2073
H2074
GENERATION OF NO FROM NITRITE BY CARBONIC ANHYDRASE
AJP-Heart Circ Physiol • VOL
38. Ridderstrale Y, Hanson M. Histochemical study of the distribution of
carbonic anhydrase in the cat brain. Acta Physiol Scand 124: 557–564,
1985.
39. Shiva S, Huang Z, Grubina R, Sun J, Ringwood LA, MacArthur PH,
Xu X, Murphy E, Darley-Usmar VM, Gladwin MT. Deoxymyoglobin
is a nitrite reductase that generates nitric oxide and regulates mitochondrial
respiration. Circ Res 100: 654 – 661, 2007.
40. Stankevicius E, Lopez-Valverde V, Rivera L, Hughes AD, Mulvany
MJ, Simonsen U. Combination of Ca2⫹-activated K⫹ channel blockers
inhibits acetylcholine-evoked nitric oxide release in rat superior mesenteric artery. Br J Pharmacol 149: 560 –572, 2006.
41. Strange RW, Dodd FE, Abraham ZH, Grossmann JG, Bruser T, Eady
RR, Smith BE, Hasnain SS. The substrate-binding site in Cu2⫹ nitrite
reductase and its similarity to Zn2⫹ carbonic anhydrase. Nat Struct Biol 2:
287–292, 1995.
42. Supuran CT. Carbonic anhydrases: novel therapeutic applications for
inhibitors and activators. Nat Rev Drug Discov 7: 168 –181, 2008.
43. Swenson ER, Teppema LJ. Prevention of acute mountain sickness by
acetazolamide: as yet an unfinished story. J Appl Physiol 102: 1305–1307,
2007.
44. Torring MS, Holmgaard K, Hessellund A, Aalkjaer C, Bek T. The
vasodilating effect of acetazolamide and dorzolamide involves mechanisms other than carbonic anhydrase inhibition. Invest Ophthalmol Vis Sci
50: 345–351, 2009.
45. van Faassen EE, Bahrami S, Feelisch M, Hogg N, Kelm M, KimShapiro DB, Kozlov AV, Li H, Lundberg JO, Mason R, Nohl H,
Rassaf T, Samouilov A, Slama-Schwok A, Shiva S, Vanin AF,
Weitzberg E, Zweier J, Gladwin MT. Nitrite as regulator of hypoxic
signaling in mammalian physiology. Med Res Rev 29: 683–741, 2009.
46. Weber RE, Fago A. Functional adaptation and its molecular basis in
vertebrate hemoglobins, neuroglobins and cytoglobins. Respir Physiol
Neurobiol 144: 141–159, 2004.
47. West JB. Respiratory Physiology. Baltimore: Williams & Wilkins, 1990.
48. Zweier JL, Wang P, Samouilov A, Kuppusamy P. Enzyme-independent
formation of nitric oxide in biological tissues. Nat Med 1: 804 – 809, 1995.
297 • DECEMBER 2009 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on June 18, 2017
26. Li H, Cui H, Kundu TK, Alzawahra W, Zweier JL. Nitric oxide
production from nitrite occurs primarily in tissues not in the blood: critical
role of xanthine oxidase and aldehyde oxidase. J Biol Chem 283: 17855–
17863, 2008.
27. Lindskog S. Structure and mechanism of carbonic anhydrase. Pharmacol
Ther 74: 1–20, 1997.
28. Logothetis NK. What we can do and what we cannot do with fMRI.
Nature 453: 869 – 878, 2008.
29. Maren TH. Use of inhibitors in physiological studies of carbonic anhydrase. Am J Physiol Renal Fluid Electrolyte Physiol 232: F291–F297,
1977.
30. Maren TH, Swenson ER. A comparative study of the kinetics of the Bohr
effect in vertebrates. J Physiol 303: 535–547, 1980.
31. Maren TH, Rayburn CS, Liddell NE. Inhibition by anions of human red
cell carbonic anhydrase B: physiological and biochemical implications.
Science 191: 469 – 472, 1976.
32. Millar TM, Stevens CR, Benjamin N, Eisenthal R, Harrison R, Blake
DR. Xanthine oxidoreductase catalyses the reduction of nitrates and nitrite
to nitric oxide under hypoxic conditions. FEBS Lett 427: 225–228, 1998.
33. Modin A, Bjorne H, Herulf M, Alving K, Weitzberg E, Lundberg JO.
Nitrite-derived nitric oxide: a possible mediator of ‘acidic-metabolic’
vasodilation. Acta Physiol Scand 171: 9 –16, 2001.
34. Mulvany MJ, Halpern W. Contractile properties of small arterial resistance vessels in spontaneously hypertensive and normotensive rats. Circ
Res 41: 19 –26, 1977.
35. Palacios-Callender M, Hollis V, Mitchison M, Frakich N, Unitt D,
Moncada S. Cytochrome c oxidase regulates endogenous nitric oxide
availability in respiring cells: a possible explanation for hypoxic vasodilation. Proc Natl Acad Sci USA 104: 18508 –18513, 2007.
36. Palmer RM, Ferrige AG, Moncada S. Nitric oxide release accounts for
the biological activity of endothelium-derived relaxing factor. Nature 327:
524 –526, 1987.
37. Petersen MG, Dewilde S, Fago A. Reactions of ferrous neuroglobin and
cytoglobin with nitrite under anaerobic conditions. J Inorg Biochem 102:
1777–1782, 2008.